Climate change and poverty in Africa - The African Journal of ...

halffacedacidicΔιαχείριση

6 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

106 εμφανίσεις

AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


24
Climate change and poverty in Africa: Mapping hotspots of vulnerability

PK THORNTON


International Livestock Research Institute (ILRI), Nairobi, Kenya
P G JONES

Waen Associates, Gwynedd, Wales, UK
T OWIYO
World Bank, Nairobi, Kenya
R L KRUSKA

International Livestock Research Institute (ILRI), Nairobi, Kenya
M HERRERO

International Livestock Research Institute (ILRI), Nairobi, Kenya
V ORINDI
International Development Research Centre (IDRC), Nairobi, Kenya
S BHADWAL

The Energy & Resources Institute (TERI), New Delhi, India
P KRISTJANSON

International Livestock Research Institute (ILRI), Nairobi, Kenya
A NOTENBAERT

International Livestock Research Institute (ILRI), Nairobi, Kenya
N BEKELE

International Livestock Research Institute (ILRI), Nairobi, Kenya
A OMOLO

International Livestock Research Institute (ILRI), Nairobi, Kenya

Abstract
Climate change and increasing climate variability threaten the attainment of the Millennium
Development Goals (MDG), and some of the worst effects on human health and agriculture
will be in sub-Saharan Africa, particularly in vulnerable regions. The relationships between
climate change and the vulnerability of resource-poor croppers and livestock keepers and their
resilience to current and future climate variability need to be better understood. This paper
describes the generation of information that combines projected climate change in agricultural
systems with vulnerability data. The results of the analysis, in terms of vulnerable people
particularly at risk for deleterious effects of climate change, are being used for impact
assessment, targeting and priority setting, to help identify locations for specific research and
adaptation activities. Given the heterogeneity in households’ access to resources, poverty
levels and ability to cope, vulnerability assessments need to be done at the sub-national level
to help improve the adaptive capacity and coping strategies of highly vulnerable households.
Keywords: Agricultural system; Vulnerability; Targeting; Impact assessment; Poverty
Résumé
Le changement climatique et la variabilité toujours croissante du climat menacent la
réalisation des Objectifs du Millénaire pour le Développement (OMD), et certains des
impacts les plus graves sur la santé humaine et l’agriculture se produiront en Afrique
subsaharienne, tout particulièrement dans les régions vulnérables. Il nous faut mieux



Corresponding author: p.thornton@cgiar.org

AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


25
comprendre la relation entre le changement climatique et et la vulnérabilité des cultivateurs
et des gardiens de bétail ayant de faibles ressources, ainsi que leur résilience vis-à-vis de la
variabilité climatique actuelle et future. Cet article met en lumière la génération
d’information regroupant des projections en matière de changement climatique dans les
systèmes agricoles et les données traitant de la vulnérabilité. Les résultats de l’analyse, en
termes de population vulnérable particulièrement exposée au risque des effets délétères du
changement climatique sont utilisés pour l’évaluation de l’impact, la mise au point du ciblage
et des priorités ; ceci afin de faciliter l’identification de lieux propices à la recherche
spécifique et aux activités d’adaptation. Étant donné l’hétérogénéité de l’accès aux
ressources des ménages, les niveaux de pauvreté et la capacité de chacun à se débrouiller, les
évaluations de la vulnérabilité doivent être effectuées à un niveau sous national pour
permettre l’amélioration de la capacité d’adaptation des ménages plus vulnérables.
Mots clés : Système agricole ; Vulnérabilité ; Ciblage ; Evaluation de l’impact ; Pauvreté

1. Introduction
The world’s climate is continuing to change at rates that are projected to be unprecedented in
recent human history. The Third Assessment Report of the Intergovernmental Panel on
Climate Change (IPCC, 2001) indicated that the global average surface temperature increased
by about 0.6°C during the 20th century. The recent Fourth Assessment Report (IPCC, 2007)
states that ‘most of the observed increase in the globally averaged temperature since the mid-
20th century is very likely due to the observed increase in anthropogenic greenhouse gas
concentrations’. The IPCC climate model projections from 2001 suggest an increase in global
average surface temperature of between 1.4 and 5.8°C from the present to 2100, the range
depending largely on the scale of fossil-fuel burning between now and then and on the
different models used. Recent modeling work indicates that the temperature increases by 2100
may be larger than those estimated in 2001 (Stainforth et al., 2005).
The impacts of climate change are likely to be highly spatially variable. At mid- to high
latitudes, crop productivity may increase slightly for local mean temperature increases of up
to 1–3°C, depending on the crop, while at lower latitudes crop productivity is projected to
decrease for even relatively small local temperature increases (1–2°C) (IPCC, 2007). In the
tropics and subtropics in general, crop yields may fall by 10–20% by 2050 because of
warming and drying, but there are places where yield losses may be much more severe (Jones
& Thornton, 2003). At the same time, developing countries are generally considered more
vulnerable to the effects of climate change than more developed countries – this is largely
attributed to a low capacity to adapt in the developing world (Thomas & Twyman, 2005). Of
the developing countries, many in Africa are seen as being the most vulnerable to climate
variability and change (Slingo et al., 2005). The challenges for development are already
considerable, and climate change is likely to add substantially to these.
Of the planet's 1.3 billion poor people, nearly 300 million are located in sub-Saharan Africa.
About 60% of these depend on livestock for some part of their livelihood (Thornton et al.,
2002; Thomas & Rangnekar, 2004). Climate change is likely to have major impacts on poor
livestock keepers and on the ecosystems goods and services on which they depend. These
impacts will include changes in the productivity of rainfed crops and forage, reduced water
availability and more widespread water shortages, and changing severity and distribution of
significant crop, livestock and human diseases. As a result, major changes can be anticipated
in livestock systems, related to livestock species mixes, crops grown, and feed resources and
feeding strategies. These changes will occur over the same period during which Africa’s
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


26
population is projected to grow from 0.9 billion people in 2005 to nearly 2 billion by 2050
(UNPD, 2007). In addition, the demand for livestock products is rising globally and will
increase significantly in the coming decades (Delgado et al., 1999) because of income shifts,
population growth, urbanization and changes in dietary preferences. While the increased
demand will probably be met mostly by increases in chicken and pig production, ruminant
populations are also likely to increase substantially, particularly in Africa. Increased demand
for livestock products will undoubtedly present opportunities for livestock keepers to intensify
production systems (Staal et al., 2001). However, it is clear that many livestock keepers in
Africa are facing a highly volatile situation characterized by rapid change.
Yet despite the critical importance of livestock to poor people and the magnitude of the
changes likely, the intersection of climate change, crop production and livestock keeping is a
neglected area of research. Little is known about how climate interacts with other drivers of
change in agricultural systems and broader development trends. The likely impacts of climate
change on the vulnerability of resource-poor croppers and livestock keepers need to be better
understood, so that resilience to current climate variability as well as to the risks associated
with longer-term climate change can be gauged, and appropriate actions taken to increase or
restore resilience where it is threatened. Given the dynamics and complexity of the systems
involved, it is clear that in general our understanding of possible local impacts is not yet at a
level where we can address them appropriately. One activity that is a prerequisite for
increased understanding of local impacts is targeting, to help identify locations for specific
research and adaptation activities. In this paper, we outline the first stage of a multi-step
targeting process. This involved a broad-brush analysis at continental level to identify areas or
‘hotspots’ that are already vulnerable and likely to suffer substantial impacts as a result of
climate change. We outline this work (which was written up in detail in Thornton et al.,
2006a), and the major results. After discussing the limitations of the analysis and how these
might be addressed through future work, we conclude by outlining the ways this type of
analysis can be used, some policy implications of the results, and the higher-resolution
analyses that are needed to refine this targeting work.
2. A continental-scale assessment of vulnerability to climate change
Work was undertaken during late 2005 and early 2006 at the behest of the UK’s Department
for International Development (DFID), to attempt some vulnerability mapping for sub-
Saharan Africa at a sub-national level that could be used to help guide the DFID’s research
resource allocation decisions as to which activities should be concentrated where. This work
was one piece of a larger set of studies commissioned by the DFID on climate change and
appropriate research for development, such as Washington et al. (2004) and Huq & Reid
(2005), that helped to clarify critical researchable issues and capacity-building needs and
opportunities. The mapping work, by contrast, was designed to throw light on where
geographically in sub-Saharan Africa research resources might be concentrated to effectively
address the issues of the poor and vulnerable in the face of inevitable climate change.
There are many different notions of what vulnerability is, and how it is related to risk and
adaptive capacity (see the reviews by Adger et al., 2004, and Vincent, 2004). O’Brien et al.
(2004) summarize two interpretations of vulnerability in the climate change literature. The
first, the ‘end point’ approach, views vulnerability as a residual of climate change impacts
minus adaptation. The second, the ‘starting point’ approach, views it as a general
characteristic generated by multiple factors and processes. Viewing vulnerability as an end
point considers that adaptations and adaptive capacity determine vulnerability, whereas
viewing it as a starting point holds that it determines adaptive capacity. Here we took a
‘starting point’ approach, viewing vulnerability to climate change as a state that is governed
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


27
not just by climate change but by multiple processes and stressors. This involves dealing with
two types of vulnerability: biophysical vulnerability – the sensitivity of the natural
environment to exposure to a hazard, and social vulnerability – the sensitivity of the human
environment to the exposure. An impact is thus a function of hazard exposure and both types
of vulnerability.
2.1 Stage 1: Hotspots of climate hazard
The first stage of the analysis was to identify those areas of sub-Saharan Africa that appear to
be particularly at risk from climate change in the coming 50 years, i.e. to identify geographic
areas where changes in temperatures and rainfall amounts and patterns etc. may be relatively
large. This was done by downscaling the outputs from various Global Circulation Models
(GCMs) and various scenarios of the future. Length of growing period (LGP) was chosen as a
proxy for agricultural impacts as it is crop-independent and a useful integrator of changes in
rainfall amounts and patterns and temperatures. We estimated changes in the length of
growing season from current conditions to 2020 and 2050 and used these changes as
indicators of climate hazard for subsequent analysis.
For looking at various scenarios of climate change to 2050, the dataset TYN SC 2.0 was used
(Mitchell et al., 2004). These monthly data cover the global land surface at a resolution of 0.5
degrees latitude and longitude for the period 2001 to 2100. There are 20 climate change
scenarios in the complete dataset, made up of permutations of five GCMs and four scenarios
that cover a wide range of economic development, fossil fuel and population growth
possibilities (the special report on emissions scenarios (SRES) – IPCC, 2000). To cut down
on the number of GCM-by-scenario combinations, we chose two GCMs with some ability to
simulate observed rainfall patterns for Africa (Liu et al., 2003; McHugh, 2005), on the basis
that this would improve confidence in the ability of these GCMs to project future conditions
under different scenarios of change, all other things being equal. The GCMs selected were
HadCM3, the UK Hadley Centre Coupled Model version 3 (Mitchell et al., 1998) and
ECHam4, the European Centre Hamburg Model version 4 (Roeckner et al., 1996). The
rainfall differences projected to 2050 by ECHam4 are relatively large, compared with those
projected by HadCM3, so these two GCMs provide a useful contrast. To reduce the number of
scenarios in the analysis, we chose scenarios A1F1 and B1, on the basis that these two cover
most of the range of projected temperature increases to 2050 (IPCC, 2001). The ‘A’ scenarios
emphasize economic growth, the ‘B’ scenarios environmental protection. The ‘1’ scenarios
assume more globalization, the ‘2’ scenarios more regionalization (IPCC, 2000). The ‘F’
scenario is a relatively fossil-fuel-intensive scenario.
The GCM output data were downscaled to a 10-arc-minute (about 18 km) grid using
WorldCLIM, a global gridded dataset of climate normals for the period 1960–1990 (Hijmans
et al., 2005) and methods based on MarkSim, a statistical weather generator (details of the
algorithms can be found in Jones & Gladkov, 2001, and Jones & Thornton, 2000). Lengths of
growing periods were calculated using methods in Jones (1987) for current and future
conditions. These results were combined with an agricultural systems classification, on the
basis that land-use options define at least part of the livelihood strategies for millions of rural
people who depend at least to some extent on natural resources for their well-being. The Seré
& Steinfeld (1996) system classification is livestock based, and to expand the classification to
include other important communities whose livelihoods are not dependent on livestock we
used the FAO farming systems classification outlined in Dixon & Gulliver (2001), which
itself is based on a principal livelihoods approach and has been used to assess general trends
in the poverty levels associated with each system in the coming decades. The classification
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


28
itself is based on FAO data and expert knowledge, and it is probably not entirely ‘mappable’
from driver variables in global- or continental-level datasets.
We created an extended systems classification by overlaying version 3 of the Seré & Steinfeld
classification (Kruska et al., 2003; Kruska, 2006) with the FAO classification, and from those
areas that were classified as ‘other’ (i.e. non-livestock systems), we used five other systems
from the FAO typology. The agricultural systems used are shown in Table 1, together with the
source of the typology and references to the mapping methods used (it should be noted that
the root-based systems make up part of the ‘other’ category). As might be expected, given the
very different ways the two classifications were derived, there are some mismatches between
them, in terms of areas that are classified inconsistently. Thus, for example, the coastal
artisanal fishing system has goats and poultry (Dixon & Gulliver, 2001), although in our
mapping of the Seré & Steinfeld system, these are classified as systems with no livestock.
Overall, however, given the continental scale of these datasets, the matching between the two
systems was found to be fairly consistent, and adequate for our purposes here.
We then overlaid the LGP changes on the agricultural systems map, to identify those systems
most at risk from both positive and negative (but mostly negative) changes in LGP. An
example of the output is shown in Figure 1, which maps the areas of Africa that are classified
as LGA and MRA systems (rangeland-based arid-semiarid, and mixed rainfed arid-semiarid,
respectively) projected to undergo at least a 20% reduction in LGP to 2050, using downscaled
outputs from the HadCM3 model for scenarios A1F1 and B1.
Table 1: Agricultural systems used in the analysis (adapted from Thornton et al., 2006a)
Code Short system description Source
COAST Coastal artisanal fishing-based systems Defined D&G
FORST Forest-based systems Defined D&G
PEREN Highland perennial-based systems Defined D&G
LGA Livestock only systems, arid-semiarid Defined S&S, mapped K
LGH Livestock only systems, humid-subhumid Defined S&S, mapped K
LGHYP Livestock only systems, hyper-arid Defined & mapped K06
LGT Livestock only systems, highland/temperate Defined S&S, mapped K
MIA Irrigated mixed crop/livestock systems, arid-subarid Defined S&S, mapped K
MIH Irrigated mixed crop/livestock systems, humid-subhumid Defined S&S, mapped K
MIHYP Irrigated mixed crop/livestock systems, hyper-arid Defined & mapped K06
MRA Rainfed mixed crop/livestock systems, arid-semiarid Defined S&S, mapped K
MRH Rainfed mixed crop/livestock systems, humid-subhumid Defined S&S, mapped K
MRHYP Rainfed mixed crop/livestock systems, hyper-arid Defined & mapped K06
MRT Rainfed mixed crop/livestock systems, highland/temperate Defined S&S, mapped K
OTHER Other systems, including root-crop-based and root-based mixed Defined S&S, D&G, mapped K
RITRE Rice-tree crop systems Defined D&G
TREEC Tree crop systems Defined D&G
URBAN Built-up areas Defined JRL
Sources: D&G: Dixon & Gulliver (2001); JRL: JRL (2005); K: Kruska et al. (2003); K06: Kruska (2006); S&S:
Seré & Steinfeld (1996)


29

































Figure 1: Areas within the LGA and MRA systems projected to undergo >20% reduction in LGP to 2050: HadCM3, A1 (left), B1 (right). LGA,
rangeland-based arid-semiarid system. MRA, mixed rainfed arid-semiarid system (from Thornton et al., 2006a)
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


30
2.2 Stage 2: Hotspots of current vulnerability
The second stage of the analysis involved characterizing sub-Saharan Africa, on the same
country-by-system basis as for the climate change impacts, in terms of a set of vulnerability
indicators. The methodological issues surrounding the choice, use and interpretation of
indicators for vulnerability assessments are discussed in Vincent (2004) and Brooks et al.
(2005), for example. Given the problems associated with using other people’s lists of
vulnerability indicators developed for different purposes in different contexts, we developed
our own set of proxy indicators, guided by the reviews and experiences of others, particularly
TERI (2003), Vincent (2004), Adger et al. (2004) and Brooks et al. (2005). To do this, a list
was compiled of possible proxy variables that could be used as vulnerability indicators. It was
judged appropriate to use a sustainable livelihoods approach as the basis for the various
indicators (Carney, 1998), and so these were grouped into the relevant asset types: human,
financial, physical, social and natural. Subsequently the list of indicators was revised in the
light of data availability issues, but the original list was left as intact as possible (indicators
relating to biodiversity and freshwater fish resources had to be omitted). Indicators of
biophysical and social vulnerability were adopted or formed from existing data sources, some
at national and some at sub-national level. Details of the 14 indicators used and data sources
are shown in Table 2, and a brief summary follows.
Table 2: Vulnerability indicators used in the analysis (from Thornton et al., 2006a)

Type Indicator Descriptor Hypothesised
functional relationship
with vulnerability
Data source
1
Natural
capital
Crop suit
Suitability
for crop
production
For all cropped
pixels, derive the
agricultural
suitability (scale 1
to 8)
The higher the
suitability, the higher
the potential crop
production, the more
potential vulnerability of
households to
substantial changes in
climate
Agricultural
suitability layer
(FAO, 2000)

GLC 2000 cropland
(JRL, 2005)
2 Natural
capital
Soil deg
Soil
degradation
due to wind,
water and
human-
induced
erosion
Four categories
(low to high) of
potential soil
degradation
The higher the soil
degradation potential,
the higher the
vulnerability
GLASOD (FAO, 2000)
3 Natural
capital
Basin
Internal
water
resources by
sub-basin
A measure of water
resources for each
pixel, from none to
high in 6 classes
The more internal water,
the lower the
vulnerability of the
household
FAO Atlas of Water
Resources and Irrigation
in Africa (FAO, 2005)
4 Physical
capital
Mkt access
Accessibility
to markets
Continuous index
based on travel
time to nearest
urban areas
The closer to the market,
the more diversified
income can be and the
higher the resilience to
shocks, even when farm
sizes are small. Better
access to markets also
implies better service
provision


Accessibility layer,
http://grid2.cr.usgs.
gov/globalpop
/africa
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


31
5 Social
capital
HPI
Human
Poverty
Index
Composite index:
probability at birth
of reaching age 40;
adult literacy rate;
% population with
no sustainable
access to improved
water source;
% children
underweight for
age
Higher HPI-1 implies
higher social capital
available
UNDP country-level data
(UNDP, 2005)
6 Social
capital
Gov
Governance Country-level data
on voice and
accountability, and
government
effectiveness

Better governance
promotes foreign
investment and creates
more jobs. A higher
index means more social
capital
World Bank composite
data (Kaufmann et al.,
2005)
7 Human
capital
Child 5
Stunting,
poverty
% children under 5
who are stunted
Stunting is one measure
of food security and a
proxy for poverty
FAO sub-national data
www.povertymap.net
8 Human
capital

Inf mort
Infant
mortality
rate, poverty

Mortality rate of
infants

Higher infant mortality
rates imply higher levels
of vulnerability

CIESIN sub-national
data,
http://beta.sedac.ciesin.
columbia.edu
9 Human
capital
Underweight
% children
underweight,
poverty

% children under 5
who are
underweight for
their age
Higher rates of
underweight children
imply higher levels of
vulnerability
CIESIN sub-national data
http://beta.sedac.ciesin.
columbia.edu
10 Human
capital
Malaria
Malaria risk Climatic suitability
for endemic
malaria
Areas with higher risk
of malaria are more
vulnerable
MARA (1998)
11 Human
capital
Pub hlth
Public health
expenditure

Public health
expenditure, as a
% of GDP
Areas are less
vulnerable with higher
government expenditure
on public health
Country-level data
(HDR, 2005)
12 Human
capital
HIV
HIV/AIDS
prevalence
Proportion of
working population
(15–49) with
HIV/AIDS
Areas with higher
prevalence of
HIV/AIDS are more
vulnerable
Country-level data
(HDR, 2005)

13 Financial
capital
GDP Ag
Agricultural
GDP
Agricultural GDP
as % of total GDP

Economies with higher
dependence on
agriculture are less
diverse and more
susceptible to climatic
events
Country level data for
2005, World Bank
http://econ.worldbank.org
14 Financial
capital
Int con
Global inter-
connectivity

The difference
between all exports
as a % of GDP and
all imports as a
% of GDP
Economies with higher
dependence on imports
are more vulnerable to
climate change and
extreme events
Country level data for
2005, World Bank
http://econ.worldbank.org
Sources: D&G: Dixon & Gulliver (2001); JRL: JRL (2005); K: Kruska et al. (2003); K06: Kruska (2006); S&S:
Seré & Steinfeld (1996)
Three indicators relate to natural capital. The first is crop suitability, representing the
suitability of different areas for crop production. This is based on a soil production index that
considers the suitability of the best adapted crop to each soil’s condition in an area and makes
a weighted average for all soils present in a cell on the basis of the characteristics of that soil.
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


32
Suitability is then ranked on a scale from 1 (least suitable) to 6 (most suitable). This is then
overlaid with a binary (absence/presence) crop distribution layer. We hypothesized that
vulnerability increases with an increase in crop suitability, as household livelihoods are more
at risk from substantial changes in climate.
The second indicator is the severity of human-induced degradation, for which we used the
data of FAO (2000). The severity of human-induced wind and water erosion is indicated by a
combination of the degree and the relative extent of the degradation process. The erosion
categories are classified into six major classes of degradation, from none to very severe. The
hypothesis here is that the higher the human-induced soil degradation potential, the higher the
vulnerability of the household.
The third indicator relating to natural capital is the extent of internal renewable water
resources (IRWR) within a sub-basin (of which there are some 600 in Africa) in mm per year.
These data are from FAO (2005) and express the difference between the natural outflow and
the natural inflow calculated by simple water-balance and hydrological models, thus
representing the sub-basin contribution to the overall runoff of the major basin. In cases where
the natural outflow is less than the natural inflow, IRWR is zero. The hypothesis is that the
more internal water available in the landscape, the lower the vulnerability of the household.
We included one indicator of physical capital, accessibility to markets. This is a continuous
index, calculated on the basis of a road network with a ‘travel time’ associated with each
stretch of road and a map of populated places. For each node on the road network,
accessibility potential is calculated based on the weighted population of the nearest populated
places on the network. The weights are based on the travel time to the nearest market centers.
The index represents the relative accessibility to markets for every pixel in the study area.
From a livelihoods perspective, the hypothesis here is that the closer a household is to the
market, the more diversified household income sources can be. The household is also likely to
have better service provision.
For social capital, we used two indicators. One is the human poverty index for developing
countries (HPI-1). This measures deprivation in the three basic dimensions of human
development captured in the Human Development Indicator (HDI):
• The probability of death at a relatively early age, measured by the probability at birth
of not surviving to age 40;
• Exclusion from the world of reading and communications, as measured by the adult
illiteracy rate;
• Lack of access to overall economic provisioning, as measured by the unweighted
average of two other indicators: the percentage of the population without sustainable
access to an improved water source, and the percentage of children underweight for
their age.
Details of HPI-1 are given in UNDP (2005), and data are at a national level. The assumption
here is of a linear inverse relationship between HPI-1 and vulnerability.
The second indicator of social capital relates to governance. Kaufmann et al. (2005) present
national indicators for six dimensions of governance: voice and accountability, political
instability and violence, government effectiveness, regulatory quality, rule of law, and control
of corruption. Their indices are based on several hundred individual variables measuring
perceptions of governance drawn from many data sources. Each indicator is normally
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


33
distributed with a mean of zero and a standard deviation of unity. These six indicators cannot
meaningfully be averaged for a particular country so, following Brooks et al. (2005), we took
two of the six (voice and accountability, and government effectiveness) and assigned the
scores to quintiles, averaged the quintile scores and then rearranged these into new quintiles.
‘Voice and accountability’ includes several indicators that measure various aspects of the
political process, civil liberties and political rights, together with the independence of the
media. ‘Government effectiveness’ combines information on the quality of public service
provision, the quality of the bureaucracy, the competence of civil servants, the independence
of the civil service from political pressures, and the credibility of the government’s
commitment to policies.
The next six indicators are related to human capital. The first three of these have been widely
used as proxies for poverty, and the hypothesis is of a direct relationship between
vulnerability and these indicators:
• The rate of chronic undernutrition, using stunting in growth among children under five
years of age as an indicator. This reflects long-term cumulative effects of inadequate
food intake and poor health conditions as a result of lack of hygiene and recurrent
illness in poor and unhealthy environments. The FAO sub-national dataset was used,
located at www.povertymap.net.
• Infant mortality, derived by dividing the number of babies who die before their first
birthday by the number of live births in that year, and multiplying by 1000. The sub-
national dataset from the Socioeconomic Data and Applications Center (SEDAC) at
CIESIN (Center for International Earth Science Information Network) was used.
• The percentage of children under five years of age who are underweight for their age.
Wasting indicates current acute malnutrition. The sub-national dataset from CIESIN
was used for this indicator.

The fourth human capital indicator used was the malaria risk (MARA, 1998). This is a
theoretical model based on available long-term climate data and shows the theoretical
suitability of local climatic conditions for stable malaria transmission in an average year. It
should be noted that malaria transmission can vary substantially from one year to the next, as
a result of climatic conditions and malaria control activities. Where the climate is suitable,
malaria is very likely to be endemic, and where it is unsuitable, malaria is likely to be
epidemic or absent. Areas with a higher risk of malaria are hypothesized to be more
vulnerable.
The fifth indicator of human capital used relates to public health expenditure. Country-level
data from UNDP (2005) were used, representing the current and capital spending on health
from central and local government budgets, external borrowing and grants, and social or
compulsory health insurance funds, expressed as a percentage of the country’s Gross
Domestic Product (GDP). The hypothesis here is that the higher the health expenditure as a
proportion of GDP, the lower the vulnerability.
The sixth indicator of human capital used was the prevalence of HIV/AIDS. Country-level
data from UNDP (2005) were used, referring to the percentage of people aged 15–49 infected
with HIV. The hypothesis here is that areas with higher rates of HIV/AIDS are more
vulnerable. HIV/AIDS is a major development issue facing sub-Saharan Africa, reducing
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


34
accessibility to labor and diminishing household productivity, among other impacts (Drimie,
2002).
There are two indicators of financial capital. One is the share of total GDP that is associated
with agriculture. Economies with a higher dependence on agriculture are hypothesized to be
less diverse and thus more susceptible to climatic events and changes. Data are at national
level and taken from UNDP (2005). The second relates to global interconnectivity, and is the
trade balance in terms of all goods and services exported and imported, expressed as a
percentage of GDP. Economies with a higher dependence on imports are hypothesized to be
more vulnerable to climate change and variability.
To help characterize the hotspots of climate hazard identified in Section 2.1 above in terms of
their vulnerability, we distilled the 14 indicators that were identified above to a smaller
number of indicators, using a Principal Components Analysis (PCA). All data were pixelized,
whatever the resolution, and PCA was carried out on all pixels that had valid data for all 14
indicators. PCA is an example of factor analysis, a class of statistical methods that attempts to
reduce the complexity of multivariate datasets by producing a set of new factors or
components that are orthogonal, thereby avoiding the problems of correlation among
indicators. A disadvantage is that the new factors may not be easily interpretable. The PCA
was done with a Varimax orthogonal rotation, and new factors were selected that had an
eigenvalue greater than unity (SAS, 1994). Before the analysis, all indicators were
transformed so that increases in their value were associated with increases in vulnerability.
The correlation matrix for the 14 indicators is shown in Table 3. All but six of these are
statistically significant at the 1% level. The relationships between governance, public health
and the poverty proxies (stunting, wasting and infant mortality) are noteworthy. Increasing
HIV/AIDS prevalence is strongly associated with increasing ‘public health vulnerability’ (i.e.
lower levels of government expenditure on public health), but apparently negatively
associated with wasting, for example. From the PCA we identified four new factors
(combinations of the 14 original indicators), and between them these four factors explained
63% of the variance in the original dataset. To derive an ‘overall’ vulnerability indicator we
calculated the weighted sum of the four components, and for weights we used the percentage
of variance explained. The resulting indicator was then normalized, and we grouped pixels
into quartiles, and aggregated the data into systems by country. For combinations of country
and system where there were missing data we used the national mean quartile across all other
systems for which there were data, as a proxy for the missing system. Human development
indicator data are not reported for Somalia in UNDP (2005), so Somalia was omitted from
this piece of the analysis. The quartiles of the resulting composite indicator are mapped in
Figure 2.
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


35

Table 3: Correlation matrix for the 14 vulnerability indicators in Table 2 for sub-
Saharan Africa (from Thornton et al., 2006a)

Malaria
Soil deg
Pub hlth
Child 5
Crop suit
GDP Ag
HPI
Inf mort
Int con
Gov
HIV
Undweight
Mkt access
Soil deg
-0.19












Pub hlth
0.06
-0.25











Child 5
-0.08
0.05
-0.52










Crop suit
0.11
0.03
-0.10
0.06









GDP Ag
-0.07
-0.11
0.78
-0.52
-0.10








HPI
0.12
-0.04
0.35
-0.27
-0.08
0.54







Inf mort
0.27
-0.01
-0.48
0.43
0.11
-0.48
-0.24






Int con
-0.14
0.01
-0.38
0.34
0.05
-0.37
-0.32
0.27





Gov
0.23
0.08
-0.42
0.24
0.11
-0.35
-0.06
0.37
0.07




HIV
-0.32
-0.18
0.58
-0.13
-0.25
0.47
0.17
-0.34
-0.04
-0.37



Undweight
0.05
0.16
-0.77
0.65
0.15
-0.64
-0.25
0.56
0.42
0.43
-0.53


Mkt access
-0.15
0.20
0.14
-0.31
-0.04
0.20
0.12
-0.24
-0.18
-0.18
0.02
-0.25

Basins
-0.12
-0.01
<0.01
0.14
-0.11
-0.11
0.16
<-0.01
0.20
-0.19
0.23
0.10
-0.09
Note: Variable codes are shown in bold in Column 2 of Table 2.

AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


36

Figure 2: Country-by-systems, showing quartiles of the vulnerability indicator derived
through PCA (quartile 1, ‘less vulnerable’ – quartile 4, ‘more vulnerable’) (from
Thornton et al., 2006a)

AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


37
2.3 Stage 3: Hotspots of vulnerability to climate change
Because of the dangers of overinterpreting the results, we did not carry out a detailed
characterization of the climate change hotspots and areas of high vulnerability. Rather, we
made a qualitative synthesis of the results, linking those systems in broad regions of Africa
that are both vulnerable and possibly subject to losses in LPG by 2050. The results are shown
in a table that divides the vulnerability/climate change space into four quadrants (Table 4).
This information was derived from inspection of the vulnerability quartiles map (Figure 2),
maps of the land-based livelihood systems tabulated in Table 1, and maps similar to Figure 1
showing the projected percentage of LGP changes by 2050 for the four different GCM-
scenario combinations. For this synthesis, the LGP changes projected by both the HadCM3
and ECHam4 GCMs were qualitatively combined.
Table 4: Synthesis of possible regions and systems affected in terms of LGP loss and
vulnerability quartile, for the A1 (top) and B1 (bottom) scenarios (both ECHam4 and
HadCM3). LGA, rangeland-based arid-semiarid system. MRA, mixed rainfed arid-
semiarid system (from Thornton et al., 2006a)
A1F1 Highest vulnerability quartile
(4)
Second-highest vulnerability
quartile (3)
Possibly severe
LGP loss (>20% to
2050)
• Some MRA systems in Sahel
• Mixed rainfed and highland
perennial systems in Great Lakes
region of E Africa
• LGA systems in parts of E Africa
• MRA, LGA systems in large parts of
Sahel
• Livestock systems and some mixed
systems in parts of E and southern
Africa
• Coastal systems in E and parts of
southern Africa
Possibly moderate
LGP loss (5–20% to
2050)
• Mixed systems in parts of E Africa

• Coastal systems of parts of W Africa
• Tree crop systems in parts of
W Africa
• Forest-based systems in central
Africa
• Root-based and root-mixed systems
in south central Africa

B1

Highest vulnerability quartile
(4)
Second-highest vulnerability
quartile (3)
Possibly severe
LGP loss (>20% to
2050)
• Some MRA systems in Sahel
• Some mixed and LGA systems in
parts of E Africa
• Scattered MRA, LGA systems in
parts of Sahel
• Livestock systems and some mixed
systems in parts of southern Africa
• Coastal systems in E and parts of
southern Africa
Possibly moderate
LGP loss (5–20% to
2050)
• Mixed rainfed systems in Great
Lakes region of E Africa
• Some MRA systems in Sahel

• Forest-based systems in central
Africa
• Livestock systems and some mixed
systems in parts of southern Africa
• Mixed systems in parts of W Africa

AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


38
Under the A1F1 scenario, there are several areas that are both in the highest vulnerability
quartile and subject to possibly severe climate change. These include some of the MRA
(mixed rainfed, arid-semiarid) systems in the Sahel, mixed rainfed systems and highland
perennial systems in the Great Lakes region of East Africa, and LGA (rangeland, arid-
semiarid) systems in parts of East Africa. Also in the highest vulnerability quartile in areas
where moderate LGP losses are possible are the mixed systems in parts of East Africa. The
areas in the second-highest vulnerability quartile that may be subject to possibly severe
climate change include the MRA and LGA systems in large parts of the Sahel, livestock
systems and some mixed systems in parts of East and southern Africa, and coastal systems in
East and parts of southern Africa. Other areas in the second highest vulnerability quartile that
may suffer moderate climate change include the coastal systems and tree-crop systems in
parts of West Africa, forest-based systems in Central Africa, and the root-based and root-
mixed systems in the southern parts of Central Africa.
The situation under the B1 scenario is qualitatively similar. While there are fewer areas in the
highest-vulnerability-severest-climate-change quadrant than for the A1F1 scenario, MRA
systems are still affected in the Sahel, as are LGA and some mixed systems in parts of East
Africa. The climate change effects on the Great Lakes region of East Africa are less than in
the A1F1 scenario, but this is still a vulnerability hot-spot. The situation is similar for the
areas in the second highest vulnerability quartile that may suffer severe climate change: the
coastal systems in East and southern Africa are still in this quadrant, as are the livestock and
some mixed systems in southern Africa particularly. The effects in the MRA and LGA
systems of the Sahel are more scattered than in the A1F1 scenario, but they are still likely to
be very important. For the moderate climate change impacts and the second highest
vulnerability quartile, the forest-based systems in Central Africa are in this quadrant, as in the
A1F1 scenario, and there are still quite large areas in the livestock and, to a lesser extent,
mixed systems in southern Africa that fall in this quadrant.

3. Discussion
There are several limitations associated with the analysis. An important one is that we are
probably underestimating the extent of climate-related hazards because no direct account is
taken of extreme events such as droughts and flooding, nor of the fact that the variability of
weather patterns in many places is increasing and with it the probability of extreme events and
natural disasters (Kasperson et al., 2005). In addition, there are various uncertainties
associated with the GCMs themselves. While the science of GCM development is continuing
to develop rapidly, different models have different capabilities for representing current (and
possible future) conditions, and there are considerable uncertainties in the science of climate
modeling itself.
Another limitation relates to the choice of vulnerability indicators. Considerably more
exploratory analysis could be undertaken with a wider range of candidate vulnerability
indicators, perhaps using other data reduction methods. Whether the broad vulnerability
groupings derived are relatively robust or not remains to be investigated, but the PCA
methodology does have the advantage of dealing with correlations between component
indicators, and it seems that this is one way also to deal with data layers at different
resolutions. It is likely that some of the richness of the component indicators is lost in the
process, however.
There are also limitations with the analysis related to its coverage. One limitation is the
treatment of coastal ecosystems, which are among the most productive yet highly threatened
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


39
systems in the world. Such systems produce disproportionately more human well-being
services than most other systems, even those covering much larger areas (Agardy & Alder,
2005). Nearly 40% of the people in the world live within 100 km of coasts, and coastal
populations are increasing rapidly – populations that are at risk from flooding and a rise in
sea-level (Nicholls, 2004), something else not considered here. Another limitation is the
absence of any treatment of fisheries and fresh-water aquaculture (we were unable to find
continental-scale data on fisheries and aquaculture issues). While the vulnerability analysis
did include indicators related to malaria risk and HIV/AIDS, there are many other potential
impacts of climate change on human health in the form of infectious diseases (Patz et al.,
2005).
Given the uncertainties and limitations, many of which warrant considerably more work, it is
likely that we have been fairly conservative in identifying hotspots. The results of the analysis
are thus indicative only. Even so, they are useful for several purposes. First, they can be used
for targeting appropriate research and adaptation interventions, to help answer questions about
where scarce research and development resources might appropriately be expended. The
analysis outlined above has been used at the DFID to help identify hotspots where climate
change related intervention activities might be concentrated. More generally, as these data
analyses are refined over time, they could provide standardized input to a broad range of
targeting work, including regional site selection for specific types of intervention (such as in
Thornton et al., 2006b, for example).
Second, these results can be used to answer more complex questions about priority setting,
such as the comparative economic analysis of a range of different interventions, to assess
likely impacts on the environment and on poverty, for instance. In the same way as for
targeting, such data could provide standardized input to regional priority setting exercises
such as that carried out by ASARECA (2005). Vulnerability information can usefully inform
the search for effective and feasible research and policy interventions, in relation to particular
characteristics of the development domains that are under consideration.
Third, the results can be used as input to formal ex ante impact assessment studies that seek to
quantify the costs and benefits of specific research-for-development activities, in terms of
likely impacts on key outputs of interest such as poverty levels and producer incomes. These
spatial vulnerability data can be used to help locate and characterize the populations that may
be affected by specific activities (which may be crop or livestock based, for instance), and
also to help quantify specific impacts. While no generic and comprehensive framework yet
exists for assessing adaptation and mitigation options for the farming households of Africa,
many of the components needed for such a framework already exist. Vulnerability data will
be a key constituent of generic impact assessment tools in the future.
In addition to highlighting various systems that may be particularly at risk, the work here has
underlined one other important message: macro-level analyses, while useful, can hide an
enormous amount of variability in what may be complex responses to climate change. There
is considerable heterogeneity in households’ access to resources, poverty levels and ability to
cope. Vulnerability and impact assessment work can certainly be usefully guided by macro-
level analyses, but ultimately this work has to be done at higher resolutions. Targeting might
thus usefully be seen as a multi-stage process, where hotspots are identified through broad-
brush analysis, followed by subsequent zooming-in to these hotspots to allow more detailed
impact assessment to be carried out at the community or household level. Such work calls for
different tools, and these might include crop, livestock and household simulation models so
that the resource, economic and household well-being implications of changes in climate and
climate variability can be appropriately assessed. Additionally, there are likely to be
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


40
significant trade-offs to consider. As a simple example, there are many areas of sub-Saharan
Africa where maize yields are likely to decrease, and these decreases may be substantial in
some areas (Jones & Thornton, 2003). In the maize-based mixed systems of sub-Saharan
Africa, maize stover is a key dry-season feed resource for cattle. A question to ask is what the
implications are of decreases in both maize grain yield and stover yield on human food
security at the household level and on livestock productivity, and what options are available to
the household to secure adequate supplies of food for household members and feed for cattle.
There is a clear dichotomy between the magnitude of the problems facing sub-Saharan Africa
on the one hand, and the necessity of helping local communities adapt in ways that fit (highly
variable) local conditions on the other. To deal with this, policy and research outputs and
interventions will have to be far better targeted in future. At the same time, there will need to
be many institutional and organizational changes that ensure that communities take centre
stage in conducting vulnerability analysis and implementation to enhance their long-term
capacities for adaptation (Yamin et al., 2005).

4. Conclusions and policy implications
Climate change poses a serious threat to development. Scholes & Biggs (2004) refer to sub-
Saharan Africa as the food crisis epicenter of the world, and conclude that projected climate
change during the first half of the 21st century will make this situation worse. Climate change
will add to the burdens of those who are already poor and vulnerable. At the same time,
agriculture in sub-Saharan Africa will continue to play a crucial role through its direct and
indirect impacts on poverty, as well as in providing an indispensable platform for wider
economic growth that reduces poverty far beyond the rural and agricultural sectors (DFID,
2005). The indicative results of this analysis show that many vulnerable regions are likely to
be adversely affected in sub-Saharan Africa, including the mixed arid-semiarid systems in the
Sahel, arid-semiarid rangeland systems in parts of East Africa, the systems in the Great Lakes
region of East Africa, the coastal regions of East Africa, and many of the drier zones of
southern Africa.
There are at least two policy implications of the work reported here. One is the seemingly
perennial problem of data, in terms of both its availability and its use. The availability of
appropriate data for carrying out vulnerability and impact analyses is a key issue, and in many
parts of Africa there are serious problems with the existing data collection systems (Lynam,
2006). It is clear that vulnerability analyses such as that outlined above make use of a wide
range of information, and Table 2 illustrates the numerous sources and scales of the
information needed. There is a continuing need for baseline data to improve targeting and
priority setting and, as Lynam (2006) and others note, considerable and widespread
collaboration is needed for data collection and utilization activities in the African context,
which are likely to require an increase in policy action.
The second policy implication is linked to the spatial heterogeneity of both farming and
livelihood systems in much of Africa and of the localized impacts of climate change. While
this certainly highlights the need for higher-resolution system studies, as noted above, it also
highlights the need to acknowledge that there may be a considerable mismatch between the
magnitude of the problems facing sub-Saharan Africa and the size of the likely development
domain for specific options for helping communities adapt that are appropriate to local
conditions. The development domains for climate change interventions may thus be
geographically relatively small. If this is indeed the case, then there are clear lessons for the
design, implementation and assessment of research and development activities designed to
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


41
address the problems brought about by a changing climate and changing climate variability.
The results of the work reported here argue strongly against large ‘magic bullet’ approaches,
and in favor of smaller, better targeted local approaches and interventions. In sum,
considerable future work is needed to refine the hotspots analysis and increase the resolution
of impact studies and thus contribute to our understanding of the issues facing millions of
people who depend on natural resources for part of their livelihood and help them adapt to
inevitable change.

Acknowledgements
We are very grateful to Simon Anderson of DFID for resources to carry out the study. We
acknowledge the inputs made by Andrew Ochieng and Brian Otiende of the African Centre
for Technology Studies (ACTS), Nairobi, Kenya, and Ms Kadambari Anantram, Ms Sreeja
Nair, Dr Vivek Kumar and Ms Ulka Kelkar of The Energy and Resources Institute (TERI)
New Delhi, India. We also thank Norbert Henninger, Timothy Mitchell, Simon Carter, Jenny
Olson, Robin Reid, Ruth Doherty, Declan Conway, John Lynam, Richard Washington and
Ade Freeman. All errors and omissions, and all views expressed, remain solely our
responsibility.

References
Adger, WN, Brooks, S, Bentham, G, Agnew, M & Eriksen, S, 2004. New indicators of
vulnerability and adaptive capacity. Technical Report No. 7, Tyndall Centre for
Climate Change Research.
Agardy, T & Alder, J, 2005. Coastal systems. Chapter 19 in Ecosystems and Human Well-
Being: Vol. 1, Current State and Trends. The Millennium Ecosystem Assessment.
http://www.millenniumassessment.org
Accessed 25 January 2008.
ASARECA (Association for Strengthening Agricultural Research in Eastern and Central
Africa), 2005. Fighting poverty, reducing hunger and enhancing resources through
regional collective action in agricultural research for development. ASARECA
Strategic Plan 2005–2015, August 2005, Entebbe, Uganda.
Brooks, N, Adger, WN & Kelly, PM, 2005. The determinants of vulnerability and adaptive
capacity at the national level and the implications for adaptation. Global
Environmental Change 15, 151–63.
Carney, D, 1998. Sustainable rural livelihoods: What contribution can we make? In Carney, D
(Ed.), Department for International Development’s Natural Resources Advisers’
Conference, July 1998. DFID, London, UK.
Delgado, C, Rosegrant, M, Steinfeld H, Ehui, S & Courbois, C, 1999. Livestock to 2020: the
next food revolution. Food, Agriculture and the Environment Discussion Paper 28.
IFPRI/FAO/ILRI, Washington, DC.
DFID (Department for International Development), 2005. DFID’s Draft Strategy for Research
on Sustainable Agriculture (SRSA) 2006–2016. http://www.dfid.gov.uk/research/srsa-
consultation.pdf
Accessed 25 January 2008
Dixon, J & Gulliver, A (with Gibbon, D), 2001. Farming systems and poverty: Improving
farmers’ livelihoods in a changing world. FAO and World Bank, Rome and
Washington, DC.
Drimie, S, 2002. The impact of HIV/AIDS on land: Case studies from Kenya, Lesotho and
South Africa. HSRC (Human Sciences Research Council), Pretoria.
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


42
FAO (Food and Agriculture Organization of the United Nations), 2000. Land resource
potential and constraints at regional and country levels. World Soil Resources Reports
90. FAO, Land and Water Development Division, Rome.
www.fao.org/ag/agl/agll/terrastat
Accessed 25 January 2008
FAO (Food and Agriculture Organization of the United Nations), 2005. Water resources and
irrigation in Africa. http://www.fao.org/nr/water/aquastat/main/index.stm
Accessed 25
January 2008
Hijmans, RJ, Cameron, SE, Parra, JL, Jones, PG & Jarvis, A, 2005. Very high resolution
interpolated climate surfaces for global land areas. International Journal of
Climatology 25, 1965–78.
Huq, S & Reid, H, 2005. Climate change and development: Consultation on key researchable
issues. IIED (International Institute for Environment and Development), London, UK.
http://www.iied.org/CC/projects/ccdevconsultation.html Accessed 25 January 2008
IPCC (Intergovernmental Panel on Climate Change), 2000. Emission scenarios: Summary for
policy makers. http://www.grida.no/climate/ipcc/spmpdf/sres-e.pdf
Accessed 25
January 2008
IPCC (Intergovernmental Panel on Climate Change), 2001. Climate Change 2001: The
Scientific Basis. Cambridge University Press, Cambridge, UK.
IPCC (Intergovernmental Panel on Climate Change), 2007. Climate Change 2007: Impacts,
adaptation and vulnerability. Summary for policy makers.
http://www.ipcc.ch/pdf/assessment-report/ar4/syr/ar4_syr_spm.pdf
Accessed 25
January 2008
Jones, PG, 1987. Current availability and deficiencies data relevant to agro-ecological studies
in the geographical area covered in IARCS. In Bunting, AH (Ed.), Agricultural
environments: Characterisation, classification and mapping. CAB International, UK,
pp. 69–82.
Jones, PG & Gladkov, A, 2001. FloraMap Version 1.01. A computer tool for predicting the
distribution of plants and other organisms in the wild. CIAT CD-ROM series, Cali,
Colombia. Centro Internacional de Agricultura Tropical. CD-ROM + Guide.
Jones, PG & Thornton, PK, 2000. MarkSim: Software to generate daily weather data for Latin
America and Africa. Agronomy Journal 93, 445–53.
Jones, PG & Thornton, PK, 2003. The potential impacts of climate change in tropical
agriculture: The case of maize in Africa and Latin America in 2055. Global
Environmental Change 13, 51–9.
JRL (Joint Research Laboratory), 2005. GLC 2000 (Global Land Cover) data layer, Joint
Research Laboratory, Ispra, Italy. http://www-gvm.jrc.it/glc2000/
Accessed 25
January 2008.
Kasperson, RE, Dow, K, Archer, ERM, Cáceres, D, Downing, TE, Elmqvist, T, Eriksen, S,
Folke, C, Han, G, Iyengar, K, Vogel, C, Wilson, KA & Ziervogel, G, 2005.
Vulnerable peoples and places. Chapter 6 in Ecosystems and human well-being:
Current state and trends. Millennium Ecosystem Assessment.
http://www.millenniumassessment.org
Accessed 25 January 2008
Kaufmann, D, Kraay, A & Mastruzzi, M, 2005. Governance matters IV: Governance
indicators for 1996–2005. World Bank Policy Research Working Paper 3630, June
2005. http://www.worldbank.org/wbi/governance/pubs/govmatters4.html
Accessed 25
January 2008
Kruska, RL, 2006. Mapped Seré & Steinfeld livestock production systems for Africa, version
3. CD-ROM, ILRI (International Livestock Research Institute), Nairobi, Kenya.
Kruska, RL, Reid, RS, Thornton, PK, Henninger, N & Kristjanson, PM, 2003. Mapping
livestock-orientated agricultural production systems for the developing world.
Agricultural Systems 77, 39–63.
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


43
Liu, P, Meehl, GA & Wu, G, 2003. Multi-model trends in the Sahara induced by increasing
CO
2
. Geophysical Research Letters 29, L015923.
Lynam, J, 2006. Climate information and agricultural development in Africa. Unpublished
Paper, Kilimo Trust, Nairobi, Kenya.
MARA (Mapping Malaria Risk in Africa), 1998. Towards an Atlas of Malaria Risk in Africa.
MARA/ARMA, Durban.
http://www.mara.org.za/trview_e.htm#Malaria%20Distribution%20Model
Accessed
25 January 2008
McHugh, MJ, 2005. Multi-model trends in East African rainfall associated with increased
CO
2
. Geophysical Research Letters 32, L01707.
Mitchell, JFB, Johns, TC & Senior, CA, 1998. Transient response to increasing greenhouse
gases using models with and without flux adjustment. Hadley Centre Technical Note
2. UK Met Office, London Road, Bracknell, UK.
Mitchell, TD, Carter, TR, Jones, PD, Hulme, M & New, M, 2004. A comprehensive set of
high-resolution grids of monthly climate for Europe and the globe: The observed
record (1901–2000) and 16 scenarios (2001–2100). Working Paper 55, Tyndall Centre
for Climate Change Research.
Nicholls, RJ, 2004. Coastal flooding and wetland loss in the 21st century: Changes under the
SRES climate and socio-economic scenarios. Global Environmental Change 14, 69–
86.
O’Brien, K, Eriksen, S, Schjolden, A & Nygaard, L, 2004. What’s in a word? Conflicting
interpretations of vulnerability in climate change research. CICERO Working Paper
2004:04, Centre for International Climate and Environmental Research, Oslo, Norway.
Patz, JA, Confalonieri, UEC, Amerasinghe, FP, Chua, KB, Daszak, P, Hyatt, AD, Molyneux,
D, Thomson, M, Yameogo, L, Lazaro, MM, Vasconcelos, P, Rubio-Palis, Y,
Campbell-Lendrum, D, Jaenisch, T, Mahamat, H, Mutero, C, Waltner-Toews, D &
Whiteman, C, 2005. Human health: ecosystem regulation of infectious diseases.
Chapter 14 in Ecosystems and Human Well-Being: Volume 1, Current State and
Trends. The Millennium Ecosystem Assessment.
http://www.millenniumassessment.org
Accessed 25 January 2008
Roeckner, E, Oberhuber, JM, Bacher, A, Christoph, M & Kirchner, I, 1996. ENSO variability
and atmospheric response in a global coupled atmosphere-ocean GCM. Climate
Dynamics 12, 737–54.
SAS, 1994. SAS/STAT User’s Guide, Volume 1. SAS Institute Inc., Cary, NC, USA.
Scholes, RJ & Biggs, R (Eds.), 2004. Ecosystem services in southern Africa: a regional
assessment. Millennium Ecosystem Assessment.
http://www.millenniumassessment.org
Accessed 25 January 2008
Seré, C & Steinfeld, H, 1996. World livestock production systems: Current status, issues and
trends. FAO Animal Production and Health Paper 127. FAO (Food and Agriculture
Organization of the United Nations), Rome, Italy.
Slingo, JM, Challinor, AJ, Hiskins, BJ & Wheeler, TR, 2005. Introduction: Food crops in a
changing climate. Philosophical Transactions of the Royal Society, Series B 360,
1983–89.
Staal, SJ, Ehui, S & Tanner, JC, 2001. Livestock–environment interactions under intensifying
production. In DR Lee, CB Barrett (Eds.), Tradeoffs or synergies? Agricultural
intensification, economic development and the environment, CAB International,
Wallingford, UK, pp. 345–64.
Stainforth, DA, Aina, T, Christensen, C, Collins, M, Faull, N, Frame, DJ, Kettleborough, JA,
Knight, S, Martin, A, Murphy, JM, Piani, C, Sexton, D, Smith, LA, Spicer, RA,
Thorpe, AJ & Allen, MR, 2005. Uncertainty in predictions of the climate response to
rising levels of greenhouse gases. Nature 433, 403–6.
AfJARE Vol 2 No 1 March 2008 PK Thornton, PG Jones, T Owiyo, RL Kruska, M Herrero, V Orindi, S Bhadwal, P Kristjanson, A Notenbaert, N Bekele and A Omolo


44
TERI (The Energy and Resources Institute) (2003). Coping with global change: vulnerability
and adaptation in Indian agriculture. The Energy and Resources Institute, New Delhi,
India.
Thomas, D & Rangnekar, D, 2004. Responding to the increasing global demand for animal
products: Implications for the livelihoods of livestock producers in developing
countries. In Owen, E, Smith, T, Steele, MA, Anderson, S, Duncan, AJ. Herrero, M,
Leaver, JD, Reynolds, CK, Richards, JI & Ku-Vera, JC (Eds), Responding to the
livestock revolution: The role of globalisation and implications for poverty alleviation.
British Society of Animal Science Publication 33, Nottingham University Press, pp.
1–35.
Thomas, DSG & Twyman, C, 2005. Equity and justice in climate change adaptation amongst
natural-resource-dependent societies. Global Environmental Change 15, 115–24.
Thornton, PK, Kruska, RL, Henninger, N, Kristjanson, PM, Reid, RS, Atieno, F , Odero, A &
Ndegwa, T, 2002. Mapping poverty and livestock in the developing world.
International Livestock Research Institute, Nairobi, Kenya.
Thornton, PK, Jones, PG, Owiyo, T, Kruska, RL, Herrero, M, Kristjanson, P, Notenbaert, A,
Bekele, N & Omolo, A, with contributions from Orindi, V, Ochieng, A, Otiende, B,
Bhadwal, S, Anantram, K, Nair, S, Kumar, V and Kelkar, U, 2006a. Mapping climate
vulnerability and poverty in Africa. Report to the Department for International
Development, ILRI, Nairobi, Kenya, May 2006.
http://www.dfid.gov.uk/research/mapping-climate.pdf
Accessed 25 January 2008
Thornton, PK, Stroud, A, Hatibu, N, Legg, C, Ly, S, Twomlow, S, Molapong, K, Notenbaert,
A, Kruska, R & von Kaufmann, R, 2006b. Site selection to test an integrated approach
to agricultural research for development: combining expert knowledge and
participatory Geographic Information System methods. International Journal of
Agricultural Sustainability 4 (1), 39–60.
UNDP (United Nations Development Programme), 2005. Human Development Report 2005.
International cooperation at a crossroads: Aid, trade and security in an unequal world.
http://hdr.undp.org/reports/global/2005/
Accessed 25 January 2008
UNPD (United Nations Population Division), 2007. Population Division of the Department of
Economic and Social Affairs of the United Nations Secretariat, World Population
Prospects: The 2006 Revision and World Urbanization Prospects: The 2005 Revision.
http://esa.un.org/unpp
Accessed 25 January 2008
Vincent, K, 2004. Creating an index of social vulnerability to climate change for Africa.
Working Paper 56, Tyndall Centre for Climate Change Research.
Washington, R, Harrison, M & Conway, D, 2004. African climate report. A report
commissioned by the UK Government to review African climate science, policy and
options for action. December 2004.
Yamin, F, Mitchell, T & Tanner, T, 2005. Linking climate adaptation: A research agenda. IDS
Bulletin 36 (4), 126–31.